Metodologia baseada em algoritmos genéticos para análise de ações de combate às perdas de água: aplicação em sistema benchmark com características do Nordeste brasileiro

Abstract

The sustainable distribution of drinking water in urban areas represents a critical challenge, especially given the high losses in supply systems caused by leaks, fraud, and operational inefficiencies. This study presents an innovative optimization methodology to reduce water losses in distribution networks, and the presented methodology is then applied to a reference network with characteristics of the state of Maranhão. The work begins with a literature review on the subject, followed by an analysis of the general characteristics of the distribution networks in the Northeast region of Brazil, particularly in the state of Maranhão. Based on this information, a benchmark model of a water distribution system was developed to represent the local conditions of the study area. Next, the author proposed, also as an original contribution, equations that correlate financial investment with the reduction of losses in various actions to combat these losses. Finally, using the data from the reference system, as well as the equations that correlate investment with loss reduction, combined with genetic algorithms, the analysis and identification of the optimal solution to combat losses in the modeled system were carried out. The results obtained support the development of more effective strategies for water resource management, as the solution presented by the proposed method was more efficient than other solutions obtained without applying the method.